CN116167250B - Machine room environment assessment method based on temperature difference weighting and time sequence algorithm - Google Patents

Machine room environment assessment method based on temperature difference weighting and time sequence algorithm Download PDF

Info

Publication number
CN116167250B
CN116167250B CN202310439078.2A CN202310439078A CN116167250B CN 116167250 B CN116167250 B CN 116167250B CN 202310439078 A CN202310439078 A CN 202310439078A CN 116167250 B CN116167250 B CN 116167250B
Authority
CN
China
Prior art keywords
temperature
data
temp
score
exponential smoothing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310439078.2A
Other languages
Chinese (zh)
Other versions
CN116167250A (en
Inventor
杨鹏
杨波
金依岩
戴伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Qunding Technology Co ltd
Original Assignee
Nanjing Qunding Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing Qunding Technology Co ltd filed Critical Nanjing Qunding Technology Co ltd
Priority to CN202310439078.2A priority Critical patent/CN116167250B/en
Publication of CN116167250A publication Critical patent/CN116167250A/en
Application granted granted Critical
Publication of CN116167250B publication Critical patent/CN116167250B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
    • G01K13/00—Thermometers specially adapted for specific purposes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/12—Timing analysis or timing optimisation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Air Conditioning Control Device (AREA)

Abstract

The invention discloses a machine room environment assessment method based on a temperature difference weighting and time sequence algorithm, which comprises the steps of collecting temperature data of a machine room through a temperature sensor; carrying out alignment pretreatment on temperature data; according to the historical performance data, intercepting temperature data from ten minutes to five minutes before the current time and from five minutes to the current time; predicting temperature data of ten minutes in the future by using an exponential smoothing model according to the historical performance data; solving temperature difference weighted data; and carrying out preliminary temperature condition evaluation, regional temperature condition risk evaluation and comprehensive temperature condition evaluation on the region where the temperature sensor is located. According to the method, all-weather temperature state evaluation can be performed on the machine room on the basis of no manpower, and the method is different from manual monitoring, the temperature state condition is quantized, the temperature of the machine room can be intuitively and accurately monitored by referring to the set interval, the labor cost is saved, and the monitoring efficiency and accuracy are improved.

Description

Machine room environment assessment method based on temperature difference weighting and time sequence algorithm
Technical Field
The invention relates to the technical field of machine room environment assessment, in particular to a machine room environment assessment method based on a temperature difference weighting and time sequence algorithm.
Background
With the rise of the Internet, the network becomes an indispensable part of people, and by 2021, the scale of Chinese netizens reaches 10.32 million, and compared with 2020, 12 months, 4296 million is increased, and the Internet popularity reaches 73.0%. Meanwhile, due to the influence of various factors, the applications such as online office, online medical treatment, online teaching and the like are kept to be increased rapidly, the number of data centers in the whole country affected by the fast increase is also increased, and the scale is also increased continuously; it is therefore important to ensure that servers in the data center room are able to function properly. In order to achieve the purpose, the key steps are that the environment temperature in the machine room is monitored, and the server is ensured to operate in a normal temperature range. Unfortunately, under the current technology, the temperature is required to be monitored and timely regulated and controlled, and can only be observed in real time by on-site operation and maintenance personnel, so that a great amount of manpower and material resources are wasted, meanwhile, timely monitoring and regulation cannot be achieved in time in timeliness, and the temperature is quite limited.
Therefore, we have designed a machine room environment assessment method based on temperature difference weighting and time series algorithm to solve the above problems.
Disclosure of Invention
The invention aims to provide a machine room environment assessment method based on a temperature difference weighting and time sequence algorithm, which is characterized in that all machine room temperature sensing data which need to be acquired for the last 24 hours are prepared, whether the temperature exceeds an alarm threshold value or not can be observed manually at present, and operation and maintenance personnel are required to be on duty continuously for 24 hours.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a machine room environment assessment method based on a temperature difference weighting and time sequence algorithm comprises the following steps:
s1, step: collecting temperature data of a machine room through a temperature sensor;
s2, step: carrying out alignment pretreatment on the temperature data to obtain clean historical performance data;
s3, step: according to the historical performance data, intercepting temperature data from 5 minutes before the current time to the current time;
s4, step: predicting temperature data for 10 minutes in the future using an exponential smoothing model based on the historical performance data;
s5, step: using the temperature data from 5 minutes before the current time to the current time and predicting the difference between the temperature data of 10 minutes and the high temperature threshold of the temperature sensor, and recording as
Figure SMS_1
The obtained difference is averaged and the maximum value is calculated, the average value is given with a weight of 0.1 and is marked as m, the maximum value is given with a weight of 0.9 and is marked as n, temperature difference weighted data from 5 minutes before the current time to the current time are finally obtained and are marked as temp, temperature difference weighted data from 10 minutes in the future are marked as predictive_temp, and the specific formulas for calculating temp and predictive_temp are as follows:
Figure SMS_2
wherein i represents the number of temperature sensations actually existing in the region;
s6, step: and carrying out preliminary temperature condition evaluation, regional temperature condition risk evaluation and comprehensive temperature condition evaluation on the region where the temperature sensor is located according to temp and prediction_temp.
Further preferably, in the step S1, the temperature sensor collects temperature data of the machine room, which refers to temperature data detected by the temperature sensor in 24 hours before the current time in the machine room, and records the temperature data.
Further preferably, in the step S2, the pre-processing of alignment is performed on the temperature data, which means that the temperature sensors are collected once every minute, and specific collection time of each temperature sensor has a slight difference, and the minimum time of all the temperature sensors every minute is aligned and taken as the collection time of the temperature sensor.
Further preferably, the temperature data of 10 minutes in the future is predicted in step S4 using an exponential smoothing model, wherein the exponential smoothing model includes a primary exponential smoothing, a secondary exponential smoothing and a tertiary exponential smoothing,
the basic formula of the exponential smoothing method is:
Figure SMS_3
in the method, in the process of the invention,
Figure SMS_4
representing a smoothed value of time t +.>
Figure SMS_5
Representing a smoothing constant, whose value range is [0,1 ]]，/>
Figure SMS_6
Representing the actual value of time t +.>
Figure SMS_7
Representing a smooth constant +.>
Figure SMS_8
A smoothed value representing time t-1;
when the historical performance data has no obvious trend change, the temperature data of 10 minutes in the future is smoothly predicted by using a primary index, and the prediction formula is as follows:
Figure SMS_9
in the method, in the process of the invention,
Figure SMS_10
representing a smooth constant +.>
Figure SMS_11
Representing the predicted value of the t+1 phase, i.e., the smoothed value of the present phase (t phase)>
Figure SMS_12
，
Figure SMS_13
Representing the current actual value, +.>
Figure SMS_14
Representing a weighted average of the current phase predictors, +.>
Figure SMS_15
Predicted value representing period t, i.e. smoothed value of period t +.>
Figure SMS_16
；
However, when the variation of the historical performance data subjected to data cleaning has a linear trend, the linear trend prediction model is established by performing primary exponential smoothing to predict that there is an obvious hysteresis deviation, so that correction is also required, the correction method is to perform secondary exponential smoothing on the basis of the primary exponential smoothing, find out the development direction and development trend of the curve by using the rule of the hysteresis deviation, and then establish the linear trend prediction model, so that the method is called a secondary exponential smoothing method, and the expression is as follows:
Figure SMS_17
in the method, in the process of the invention,
Figure SMS_18
a quadratic exponential smoothing value representing the t-th period, is->
Figure SMS_19
Representing the weighting coefficients (also called smoothing constants),
Figure SMS_20
an exponential smoothing value representing the t-th period, < >>
Figure SMS_21
A weighted average of the secondary exponential smoothing values representing the t-1 th period;
if the variation of the time sequence shows a quadratic curve trend, predicting by adopting a third exponential smoothing method, wherein the third exponential smoothing is performed again on the basis of the second exponential smoothing, and the calculation formula is as follows:
Figure SMS_22
in the method, in the process of the invention,
Figure SMS_23
three exponential smoothing values representing the t-th period, respectively>
Figure SMS_24
Representing the weighting coefficients (also called smoothing constants),
Figure SMS_25
a quadratic exponential smoothing value representing the t-th period, is->
Figure SMS_26
Representing a weighted average of the three exponentially smoothed values for the t-1 th period.
Further preferably, in the step S6, the preliminary temperature condition evaluation includes the following: the temperature data from the current time 5 minutes before the current time to the current time and the temperature data predicted for 10 minutes in the future are subjected to temperature difference weighting respectively to obtain different temperature difference weighted data, temp and prediction_temp, and a preliminary environment evaluation score of the area where the temperature sensor in the current machine room is located is judged according to the temperature difference weighted data, and the judgment method is as follows:
(1) When temp >0.5 or temp >0 and simultaneously predict_temp >0, then preliminary context evaluation score=temp;
(2) When temp < -0.5, then preliminary environmental assessment score=temp+0.5;
(3) When prediction_temp < -1, then preliminary context assessment score=max [ -1,0.8 ] (prediction_temp+1) ];
(4) When prediction_temp >0.5, then preliminary environmental assessment score=min [1,0.8 (prediction_temp-0.5) ];
(5) When none of the above conditions is satisfied, then the preliminary environmental assessment score=0;
further preferably, in the step S6, the regional temperature condition risk assessment includes the following:
1) Counting the times of high-temperature alarms of each temperature sensor in the past 24 hours under the current time, if no high-temperature alarms occur, considering that the area has no alarm risk, and giving an area temperature condition risk assessment risk_score=0;
2) If the temperature difference weighted data of a certain temperature sensor in the machine room is more than 1 within 2 hours, the area is considered to have alarm risk, and the area temperature condition risk assessment risk_score=1 is given;
3) If no alarm condition occurs in the last 2 hours, counting the number of continuous segments with the temperature difference weighted data of the temperature sensor being more than 0 in the last 24 hours under the current time, and recording as cnt;
total number of data segments all _ cnt = number of temperature sensors/30,
giving the regional temperature status risk assessment risk_score=cnt a preliminary environmental assessment score/all_cnt.
Further preferably, in the step S6, the comprehensive evaluation of the temperature condition includes the following:
(a) When the preliminary environmental assessment score is not less than 0, the temperature condition comprehensive assessment score=the preliminary environmental assessment score;
(b) When the preliminary environmental assessment score is less than 0, that is, the current regional temperature condition is good, if the regional temperature condition risk assessment risk_score=1, the temperature condition comprehensive assessment score=0;
(c) When the preliminary environmental assessment score < 0, if the regional temperature condition risk assessment score_score is not equal to 1, the temperature condition comprehensive assessment score=preliminary environmental assessment score (1-regional temperature condition risk assessment score_score).
Compared with the prior art, the invention has the beneficial effects that: according to the method, the temperature sensing temperature is predicted by using a time sequence algorithm, and a temperature difference weighting method is combined, so that a numerical value, namely a temperature sensing alarm condition, is obtained, the condition that operation and maintenance personnel need to keep on the whole in the initial stage of project on-line is avoided, meanwhile, the condition that the temperature sensing needs to be focused can be displayed more intuitively and accurately, the machine room can be subjected to 24-hour uninterrupted temperature state assessment on the basis of no manpower, and the method is different from manual monitoring.
Drawings
Fig. 1 is a schematic flow chart of a machine room environment assessment method based on a temperature difference weighting and time sequence algorithm.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments.
Examples
Referring to fig. 1, according to the machine room environment assessment method based on the temperature difference weighting and time sequence algorithm provided by the embodiment, the temperature sensing temperature is predicted by using the time sequence algorithm, and a numerical value, namely, a temperature sensing alarm condition is obtained by combining the temperature difference weighting method, so that the condition that operation and maintenance personnel need to keep on the whole in the initial stage of on-line of a project is avoided, meanwhile, the condition that the temperature sensing needs to be focused more intuitively and accurately can be displayed, and different air conditioner brands have certain universality for different machine rooms, and the machine room can be assessed for 24 hours of uninterrupted temperature states on the basis of no manpower. Specifically, the machine room environment assessment method comprises the following steps:
s1, step: collecting temperature data of a machine room through a temperature sensor; the temperature data of the machine room collected by the temperature sensor refer to the temperature data detected by the temperature sensor in 24 hours before the current time in the machine room, and the temperature data are recorded.
S2, step: carrying out alignment pretreatment on the temperature data to obtain clean historical performance data; the temperature data is subjected to alignment pretreatment, namely the temperature sensors are collected once every minute, the specific collection time of each temperature sensor has a slight difference, and the minimum time of all the temperature sensors every minute is aligned and taken as the collection time of the temperature sensors.
S3, step: according to the historical performance data, intercepting temperature data from 5 minutes before the current time to the current time;
s4, step: predicting temperature data for 10 minutes in the future using an exponential smoothing model based on the historical performance data;
the exponential smoothing model includes primary exponential smoothing, secondary exponential smoothing and tertiary exponential smoothing,
the basic formula of the exponential smoothing method is:
Figure SMS_27
in the method, in the process of the invention,
Figure SMS_28
representing a smoothed value of time t +.>
Figure SMS_29
Representing a smoothing constant, whose value range is [0,1 ]]，/>
Figure SMS_30
Representing the actual value of time t +.>
Figure SMS_31
Representing a smooth constant +.>
Figure SMS_32
A smoothed value representing time t-1;
when the historical performance data has no obvious trend change, the temperature data of 10 minutes in the future is smoothly predicted by using a primary index, and the prediction formula is as follows:
Figure SMS_33
in the method, in the process of the invention,
Figure SMS_34
representing a smooth constant +.>
Figure SMS_35
Representing the predicted value of the t+1 phase, i.e., the present phase (t phase)Smooth value of +.>
Figure SMS_36
，
Figure SMS_37
Representing the current actual value, +.>
Figure SMS_38
Representing a weighted average of the current phase predictors, +.>
Figure SMS_39
Predicted value representing period t, i.e. smoothed value of period t +.>
Figure SMS_40
；
However, when the variation of the historical performance data cleaned by the data has a linear trend, the linear trend prediction model is established by performing primary exponential smoothing to predict that there is an obvious hysteresis deviation, so that the correction is also needed, the correction method is to perform secondary exponential smoothing on the basis of the primary exponential smoothing, find out the development direction and development trend of the curve by using the rule of the hysteresis deviation, and then establish the linear trend prediction model, so that the linear trend prediction model is called a secondary exponential smoothing method, and the expression is:
Figure SMS_41
in the method, in the process of the invention,
Figure SMS_42
a quadratic exponential smoothing value representing the t-th period, is->
Figure SMS_43
Represents a weighting coefficient (also called smoothing constant),>
Figure SMS_44
an exponential smoothing value representing the t-th period, < >>
Figure SMS_45
Secondary index representing period t-1A weighted average of the smoothed values;
if the variation of the time sequence shows a quadratic curve trend, predicting by adopting a third exponential smoothing method, wherein the third exponential smoothing is performed again on the basis of the second exponential smoothing, and the calculation formula is as follows:
Figure SMS_46
in the method, in the process of the invention,
Figure SMS_47
three exponential smoothing values representing the t-th period, respectively>
Figure SMS_48
Representing the weighting coefficients (also called smoothing constants),
Figure SMS_49
a quadratic exponential smoothing value representing the t-th period, is->
Figure SMS_50
Representing a weighted average of the three exponentially smoothed values for the t-1 th period.
S5, step: using the temperature data from 5 minutes before the current time to the current time and predicting the difference between the temperature data of 10 minutes and the high temperature threshold of the temperature sensor, and recording as
Figure SMS_51
The obtained difference is averaged and the maximum value is calculated, the average value is given with a weight of 0.1 and is marked as m, the maximum value is given with a weight of 0.9 and is marked as n, temperature difference weighted data from 5 minutes before the current time to the current time are finally obtained and are marked as temp, temperature difference weighted data from 10 minutes in the future are marked as predictive_temp, and the specific formulas for calculating temp and predictive_temp are as follows:
Figure SMS_52
wherein i represents the number of temperature sensations actually existing in the region;
s6, step: and carrying out preliminary temperature condition evaluation, regional temperature condition risk evaluation and comprehensive temperature condition evaluation on the region where the temperature sensor is located according to temp and prediction_temp.
Wherein the preliminary temperature condition assessment includes the following: the temperature data from the current time 5 minutes before the current time to the current time and the temperature data predicted for 10 minutes in the future are subjected to temperature difference weighting respectively to obtain different temperature difference weighted data, temp and prediction_temp, and a preliminary environment evaluation score of the area where the temperature sensor in the current machine room is located is judged according to the temperature difference weighted data, and the judgment method is as follows:
(1) When temp >0.5 or temp >0 and simultaneously predict_temp >0, then preliminary context evaluation score=temp;
(2) When temp < -0.5, then preliminary environmental assessment score=temp+0.5;
(3) When prediction_temp < -1, then preliminary context assessment score=max [ -1,0.8 ] (prediction_temp+1) ];
(4) When prediction_temp >0.5, then preliminary environmental assessment score=min [1,0.8 (prediction_temp-0.5) ];
(5) When none of the above conditions is satisfied, then the preliminary environmental assessment score=0.
Wherein the regional temperature condition risk assessment includes the following:
1) Counting the times of high-temperature alarms of each temperature sensor in the past 24 hours under the current time, if no high-temperature alarms occur, considering that the area has no alarm risk, and giving an area temperature condition risk assessment risk_score=0;
2) If the temperature difference weighted data of a certain temperature sensor in the machine room is more than 1 within 2 hours, the area is considered to have alarm risk, and the area temperature condition risk assessment risk_score=1 is given;
3) If no alarm condition occurs in the last 2 hours, counting the number of continuous segments with the temperature difference weighted data of the temperature sensor being more than 0 in the last 24 hours under the current time, and recording as cnt;
total number of data segments all _ cnt = number of temperature sensors/30,
giving the regional temperature status risk assessment risk_score=cnt a preliminary environmental assessment score/all_cnt.
Wherein, the comprehensive evaluation of the temperature condition comprises the following contents:
(a) When the preliminary environmental assessment score is not less than 0, the temperature condition comprehensive assessment score=the preliminary environmental assessment score;
(b) When the preliminary environmental assessment score is less than 0, that is, the current regional temperature condition is good, if the regional temperature condition risk assessment risk_score=1, the temperature condition comprehensive assessment score=0;
(c) When the preliminary environmental assessment score < 0, if the regional temperature condition risk assessment score_score is not equal to 1, the temperature condition comprehensive assessment score=preliminary environmental assessment score (1-regional temperature condition risk assessment score_score).
The machine room environment assessment method based on the temperature difference weighting and time sequence algorithm provided by the embodiment is different from a traditional manual monitoring mode, the temperature state condition is quantized, the machine room temperature can be intuitively and accurately monitored by referring to a set interval, the labor cost is saved, the monitoring efficiency and accuracy are improved, timely monitoring and regulation are achieved in timeliness, and all the machine rooms are regulated and controlled at one time.
The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.

Claims (4)

1. The machine room environment assessment method based on the temperature difference weighting and time sequence algorithm is characterized by comprising the following steps of:
s1, step: collecting temperature data of a machine room through a temperature sensor;
s2, step: carrying out alignment pretreatment on the temperature data to obtain clean historical performance data;
s3, step: according to the historical performance data, intercepting temperature data from 5 minutes before the current time to the current time;
s4, step: predicting temperature data for 10 minutes in the future using an exponential smoothing model based on the historical performance data;
s5, step: temperature data from 5 minutes before the current time to the current time and temperature data predicted to be 10 minutes in future are respectively used for a plurality of temperature sensations existing in the region, the difference is made between the temperature data and a high temperature threshold value of a temperature sensor, the maximum value is taken, and the maximum value is recorded as x i X for several temperature senses i Summing and averaging to obtain a maximum value, wherein the average value is given with a weight of 0.1 and the maximum value is given with a weight of 0.9 and is given with a weight of n, so that temperature difference weighted data from 5 minutes before the current time to the current time are finally obtained, the temperature difference weighted data are marked as temp and 10 minutes in the future, the temperature difference weighted data are marked as predictive_temp, and formulas for calculating temp and predictive_temp are as follows:
Figure FDA0004271766790000011
wherein i represents the number of temperature sensations actually existing in the region;
s6, step: according to temp and prediction_temp, carrying out preliminary temperature condition assessment, regional temperature condition risk assessment and comprehensive temperature condition assessment on the region where the temperature sensor is located;
the preliminary temperature condition assessment includes the following: according to temp and prediction_temp, judging a preliminary environment evaluation score of the area where the temperature sensor in the current machine room is located, wherein the judging method is as follows:
(1) When temp >0.5 or temp >0 and simultaneously predict_temp >0, then preliminary context evaluation score=temp;
(2) When temp < -0.5, then preliminary environmental assessment score=temp+0.5;
(3) When prediction_temp < -1, then preliminary context assessment score=max [ -1,0.8 ] (prediction_temp+1) ];
(4) When prediction_temp >0.5, then preliminary environmental assessment score=min [1,0.8 (prediction_temp-0.5) ];
(5) When none of the above conditions is satisfied, then the preliminary environmental assessment score=0;
the regional temperature condition risk assessment includes the following:
1) Counting the times of high-temperature alarms of each temperature sensor in the past 24 hours under the current time, if no high-temperature alarms occur, considering that the area has no alarm risk, and giving an area temperature condition risk assessment risk_score=0;
2) If the temperature difference weighted data of a certain temperature sensor in the machine room is more than 1 within 2 hours, the area is considered to have alarm risk, and the area temperature condition risk assessment risk_score=1 is given;
3) If no alarm condition occurs in the last 2 hours, counting the number of continuous segments with the temperature difference weighted data of the temperature sensor being more than 0 in the last 24 hours under the current time, and recording as cnt;
total number of data segments all _ cnt = number of temperature sensors/30,
assigning a temperature condition risk assessment risk_score=cnt to the region a preliminary environmental assessment score/all_cnt;
the comprehensive evaluation of the temperature condition comprises the following contents:
(a) When the preliminary environmental assessment score is not less than 0, the temperature condition comprehensive assessment score=the preliminary environmental assessment score;
(b) When the preliminary environmental assessment score is less than 0, that is, the current regional temperature condition is good, if the regional temperature condition risk assessment risk_score=1, the temperature condition comprehensive assessment score=0;
(c) When the preliminary environmental assessment score < 0, if the regional temperature condition risk assessment score_score is not equal to 1, the temperature condition comprehensive assessment score=preliminary environmental assessment score (1-regional temperature condition risk assessment score_score).
2. The machine room environment assessment method based on the temperature difference weighting and time series algorithm according to claim 1, wherein in the step S1, the temperature sensor collects temperature data of the machine room, and refers to all temperature data detected by the temperature sensor within 24 hours from the current time in the machine room, and records the temperature data.
3. The machine room environment assessment method based on the temperature difference weighting and time series algorithm according to claim 1, wherein in the step S2, the alignment pretreatment is performed on temperature data, namely, the temperature sensors are collected once every minute, specific collection time of each temperature sensor has a slight difference, and the minimum time of all the temperature sensors every minute is aligned and taken as the collection time of the temperature sensors.
4. The machine room environment assessment method based on the temperature difference weighted sum time series algorithm according to claim 1, wherein the temperature data of 10 minutes in the future is predicted in step S4 using an exponential smoothing model, wherein the exponential smoothing model includes a primary exponential smoothing, a secondary exponential smoothing and a tertiary exponential smoothing,
the basic formula of the exponential smoothing method is:
S t ＝α*y t +(1-α)*S t-1
wherein S is t A smooth value of time t, alpha represents a smooth constant, and the value range is 0,1]，y t Representing the actual value of time t, 1-alpha representing the smoothing constant, S t-1 A smoothed value representing time t-1;
when the historical performance data has no obvious trend change, the temperature data of 10 minutes in the future is smoothly predicted by using a primary index, and the prediction formula is as follows:
y (t+1)′ ＝α*y t +(1-α)*y t
wherein α represents a smoothing constant, y (t+1)′ Representing the predicted value of the t+1 phase, alpha × y t Represents the actual value of the current period, (1-alpha) x y t′ Representing a weighted average of the current-period predictors, y t′ A predicted value representing the t period;
when the variation of the historical performance data has a linear trend, the linear trend is predicted by using primary exponential smoothing, and the linear trend is required to be corrected, wherein the correction method is to perform secondary exponential smoothing on the basis of the primary exponential smoothing, find out the development direction and the development trend of a curve by using the law of the hysteresis deviation, and then establish a linear trend prediction model, so the linear trend prediction model is called a secondary exponential smoothing method, and the expression is as follows:
Figure FDA0004271766790000041
in the method, in the process of the invention,
Figure FDA0004271766790000042
a quadratic exponential smoothing value representing the t-th period, alpha representing a smoothing constant,/for>
Figure FDA0004271766790000043
An exponential smoothing value representing the t-th period, < >>
Figure FDA0004271766790000044
A weighted average of the secondary exponential smoothing values representing the t-1 th period;
if the variation of the time sequence shows a quadratic curve trend, predicting by adopting a third exponential smoothing method, wherein the third exponential smoothing is performed again on the basis of the second exponential smoothing, and the calculation formula is as follows:
Figure FDA0004271766790000051
in the method, in the process of the invention,
Figure FDA0004271766790000052
three exponential smoothing values representing period t, alpha representing the smoothing constant, +.>
Figure FDA0004271766790000053
A quadratic exponential smoothing value representing the t-th period, is->
Figure FDA0004271766790000054
Representing a weighted average of the three exponentially smoothed values for the t-1 th period.
CN202310439078.2A 2023-04-23 2023-04-23 Machine room environment assessment method based on temperature difference weighting and time sequence algorithm Active CN116167250B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310439078.2A CN116167250B (en) 2023-04-23 2023-04-23 Machine room environment assessment method based on temperature difference weighting and time sequence algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310439078.2A CN116167250B (en) 2023-04-23 2023-04-23 Machine room environment assessment method based on temperature difference weighting and time sequence algorithm

Publications (2)

Publication Number Publication Date
CN116167250A CN116167250A (en) 2023-05-26
CN116167250B true CN116167250B (en) 2023-07-07

Family

ID=86422247

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310439078.2A Active CN116167250B (en) 2023-04-23 2023-04-23 Machine room environment assessment method based on temperature difference weighting and time sequence algorithm

Country Status (1)

Country Link
CN (1) CN116167250B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117804639B (en) * 2024-02-29 2024-05-17 潍坊盛品印刷设备有限公司 Temperature calibration method and system for temperature control sensor of cementing machine
CN117873221B (en) * 2024-03-12 2024-05-28 广州中科医疗美容仪器有限公司 Temperature monitoring control method and system for moxibustion therapy bin
CN118014313B (en) * 2024-04-08 2024-06-21 福建智联万物科技有限公司 Remote monitoring system is equipped in storage based on thing networking

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110414155B (en) * 2019-07-31 2022-09-30 北京天泽智云科技有限公司 Fan component temperature abnormity detection and alarm method with single measuring point
CN112820414B (en) * 2021-01-29 2021-11-09 南威软件股份有限公司 Early warning method for new crown epidemic situation based on improved cubic exponential smoothing model and LSTM model
CN114841410A (en) * 2022-03-31 2022-08-02 大连海心信息工程有限公司 Heat exchange station load prediction method and system based on combination strategy

Also Published As

Publication number Publication date
CN116167250A (en) 2023-05-26

Similar Documents

Publication Publication Date Title
CN116167250B (en) Machine room environment assessment method based on temperature difference weighting and time sequence algorithm
Eini et al. Smart building management system: Performance specifications and design requirements
Wang et al. Online model-based fault detection and diagnosis strategy for VAV air handling units
TWI267012B (en) Quality prognostics system and method for manufacturing processes
JP4661640B2 (en) Air conditioning control system
CN108038044B (en) Anomaly detection method for continuous monitored object
US11585549B1 (en) Thermal modeling technology
JP4386748B2 (en) Air conditioning load prediction method, air conditioning load prediction device, air conditioning load prediction program, and recording medium
US20190195525A1 (en) Method and apparatus for operating heating and cooling equipment via a network
Horrigan et al. A statistically-based fault detection approach for environmental and energy management in buildings
CN113868953B (en) Multi-unit operation optimization method, device and system in industrial system and storage medium
CN110989044B (en) Air quality index level probability forecasting method, device, equipment and storage medium
CN112365056A (en) Electrical load joint prediction method and device, terminal and storage medium
CN113036913A (en) Method and device for monitoring state of comprehensive energy equipment
CN115423301A (en) Intelligent electric power energy management and control method, device and system based on Internet of things
CN117010638B (en) Intelligent management method and system for hotel equipment
Yang et al. Toward Machine Learning-based Prognostics for Heating Ventilation and Air-Conditioning Systems.
CN111582588B (en) Building energy consumption prediction method based on triple convolution fusion GRU
CN117200223A (en) Day-ahead power load prediction method and device
JP2005122517A (en) Energy demand prediction method, energy demand prediction device and energy demand prediction program and recording medium
CN111859783B (en) Water pressure prediction method, system, storage medium, equipment and urban water supply system
CN108921340B (en) Flood probability forecasting method based on error transfer density function
JP3707589B2 (en) Electricity demand forecast method
Hou et al. Cooling load prediction based on the combination of rough set theory and support vector machine
CN117474710B (en) Hollow glass whole-process informationized management system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant